(PS5-D78) Effectiveness of Youth Psychotherapy Interventions Delivered in Schools: A Meta-analysis of Randomized Controlled Trials

Introduction: In the US, roughly 25% of young people experiences severe impairment as a result of mental illness (Merikangas et al., 2010). That said, only one in three young people affected by mental illness receives treatment (Kazdin, 2017; Kessler et al., 2005; Merikangas et al., 2011; World Health Organization, 2018). Schools are common and accessible settings for youth mental health interventions (Farmer et al., 2003; Leaf et al., 1996; Merikangas et al., 2011; Rones & Hoagwood, 1986). Schools are sometimes the only settings where ethnic minority youth, low-income youth, and immigrant youth have access to mental health services (Coleman, 2004; Yeh, 2004). This poster reports on quantitative outcome measures from a meta-analysis of youth psychotherapy interventions delivered in school settings. There are three aims: (1) describe characteristics of the mental health interventions that are currently offered in schools; (2) estimate the overall effectiveness of school-based interventions; and (3) determine which characteristics of the interventions, schools, and participants moderate treatment effectiveness.

Methods: We conducted a systematic search of the PsycInfo and PubMed databases for randomized controlled trials (RCTs) of psychotherapy interventions delivered in schools for youths with elevated symptoms of internalizing, externalizing, and multiple problems. We identified 122 studies and 207 treatment-control comparisons. Data about intervention characteristics and outcomes, schools, and participants, were used to produce descriptive statistics, treatment effectiveness estimates, and tests of candidate moderators.

Results: The mean age of participants across studies was 10.9. 60.0% of the selected studies had 50% or more male participants. We ran a robust variance estimation using the robumeta package in R. Based on preliminary analyses, school-based psychotherapy interventions have a pooled effect size of .30 (90%CI .17-.43; p< .001) at post-treatment and .14 (90%CI .02-.26; p=.02) at follow-up. Full results, including descriptive statistics and moderator analyses will be presented in the final poster.

Discussion: The pooled effect size for psychotherapy interventions delivered in schools is smaller than the pooled effect size for all youth psychotherapy interventions, which is .50 (95%CI .45-.55, p < .001) (Weisz et al., 2019). Future research could investigate potential reasons why youth psychotherapy delivered in schools may not be as effective as interventions delivered in other settings. The findings of this meta-analysis may have the potential to inform school-based mental healthcare research and practice.